The objective of this study is to clarify the individual- and community-level mechanisms by which a woman's empowerment is associated with her risk of experiencing intimate partner violence (IPV) in rural Bangladesh. It employs four innovative methodological strategies to achieve this objective. First, much of the existing research on women's empowerment has used relatively few items to capture a narrow range of dimensions of this concept. This project builds on a prior NIH developmental grant on women's empowerment in Bangladesh (#1R21HD053580 - 01 PI Schuler) by incorporating a comprehensive and contextually appropriate set of indicators of women's empowerment. Second, prior research on attitudes and norms about IPV has used a narrow range of items to capture these concepts, and such items may have been variously (mis)understood by the survey respondents. Here, new measures of attitudes regarding IPV in rural Bangladesh, developed through another developmental research grant from NIH (#1R21HD058173-01 PIs Schuler and Yount) are used. Third, prior studies of contextual influences on IPV have suffered from limited measures of both IPV and community context. This project uses a comprehensive set of measures to characterize the gendered social context, which will appropriately situate individual-level processes of women's empowerment. It also includes a comprehensive set of measures that capture the type and timing of IPV that women experience. Finally, the project combines the strengths of quantitative and qualitative methods to enhance interpretation of the mechanisms by which individual-, couple- and community-level variables operate and interact to influence a woman's risk of experiencing IPV. The quantitative data collection will combine a three-wave (a baseline plus two follow-ups) panel survey of 3500 recently married women (RMW) with one cross-sectional survey, administered at baseline, of 1500 ever-married men(EMM) and 1500 ever-married women (EMW). The survey data elicited from RWM, EMW, and EMM will make it possible to generate 3 categories of indicators to test the study hypotheses in a multilevel analytic framework: independent variables at both the village level and individual levels, individual- level IPV outcomes, and control variables. Qualitative data will consist of 1) life history narratives of young married women with relatively high levels of empowerment in which the evolution of empowerment and IPV in their lives are simultaneously elicited; and 2) group discussions with young married men, young married women (married 2-6 years), and older men and women.